Min Max Generalization for Deterministic Batch Mode Reinforcement Learning: Relaxation Schemes
Raphael FonteneauDamien ErnstBernard BoigelotQuentin LouveauxPublished in: CoRR (2012)
Keyphrases
- min max
- batch mode
- reinforcement learning
- incremental learning
- active learning
- batch mode active learning
- learning algorithm
- supervised learning
- control policy
- semi supervised
- computationally expensive
- multiple instance learning
- state space
- markov decision processes
- learning problems
- linear classifiers
- online algorithms
- dynamic programming
- semi supervised learning
- lower bound
- support vector
- multi agent